Alzheimer's Disease (AD) is the most common cause of dementia. Many AD patients develop sleep/wake cycle disturbances. We propose a randomized controlled clinical trial to carefully evaluate a nonpharmacologic approach to these problems. Our specific aims and hypotheses are: 1)Efficacy. We plan to compare the short-term efficacy of four treatments for sleep/wake cycle disturbances in community-dwelling AD patients: 1) Morning bright light treatment (AM LIGHT), 2) Evening bright light treatment (PM LIGHT), 3) Morning and Evening bright light treatment (AM&PM LIGHT) and 4) a control treatment, morning dim light treatment (AM DIM). Hypothesis 1: We hypothesize that 4 weeks of either AM LIGHT, PM LIGHT or AM&PM LIGHT treatment will be more efficacious than the AM DIM treatment in consolidating nighttime sleep as assessed by actigraphy. We do not expect to see overall differences in efficacy between AM LIGHT, PM LIGHT and AM&PM LIGHT. 2) Individual Differences in Response (Moderators). We will conduct preliminary "moderator analyses" to determine in which patients which LIGHT treatment works best. Hypothesis 2: Primary Predictors of Response: Response to the LIGHT treatments will depend upon the initial MMSE score. Secondary Predictor Analyses: We expect that baseline sleep/wake and circadian parameters and age may predict response to AM LIGHT versus PM light. 3) Effectiveness and Quality of Life. We will determine if bright light therapy is effective in terms of improvement in quality of life. Hypothesis 3: We hypothesize that any of the three LIGHT treatments will be more effective than AM DIM in improving quality of life in patients and caregivers. These hypotheses will be tested in a parallel groups design study in community-dwelling AD patients with caregivers. Specifically, we propose to randomize 120 patient/caregiver dyads into four groups of 30 dyads each: AM LIGHT, PM LIGHT AM&PM LIGHT or AM DIM. After 4 weeks of treatment (Phase 1) all subjects will be given bright light for an additional 4 weeks (Phase 2). Outcome will be documented by objective circadian activity rhythm parameters and standard sleep/wake measures both assessed by actigraphy.